Prognostic Evaluation Based on Dual-Time ¹⁸F-FDG PET/CT Radiomics Features in Patients with Locally Advanced Pancreatic Cancer Treated by Stereotactic Body Radiation Therapy

Table 2

Radiomics features were selected via the elastic net and the corresponding weights in retraining the Cox regression model.

Feature name	Categories	Modality	Time	Weights
				Early	Delay	Dual
Solidity	Shape	—	Early	−3.7884	—	−2.0451
Contrast	GLDS (HHH)	CT		0.3654		1.4669
Contrast	GLDS (HHH)	PET		0.5546		1.9619
Energy	GLDS (LLL)	CT		6.4719		—
Contrast	GLDS (HHH)	PET		−0.1456		—
Energy	GLCM (LLH-LHL-HLL)	CT		−0.1552		—
Busyness	NGTDM (LLL)	PET	Delay	—	3.5622	2.3483
Entropy	Histogram (original)	CT			0.8987	2.5573
Gray level nonuniformity	GLSZM (LLH-LHL-HLL)	CT			3.1435	3.1575
Busyness	NGTDM (original)	PET			−1.2738	—
Entropy	GLDS (original)	CT			0.3695	—
Mean	GLDS (LLL)	PET			−0.2466	—

GLCM, gray-level co-occurrence matrix; GLDS, gray-level difference statistics; GLRLM, gray-level run length matrix; GLZSM, gray-level zone size matrix; NGTDM, neighborhood gray-tone difference matrix; LHH = lowpass filter + highpass filter + highpass filter.